608 Recent Literature. [q^ 



lems involved in a study of the bird life of the group. We here learn 

 that of the 217 species recorded from the islands, 75 are regular breeders, 

 while 142 are transients or of casual occurrence. They are further grouped 

 (with some duplication) as Residents 61; Partial Residents (i. e., the 

 resident population augmented at certain seasons by migrants from else- 

 where) 5; Summer Visitors (nesting regularly but not wintering) 9; Winter 

 Visitors 15; Birds of Passage 32; Annual Visitors (time of occurrence ir- 

 regular) 5; Occasional Visitors 30; Rare Visitors 72. There are also given 

 in an appendix 25 species recorded from the islands on evidence insufficient 

 to include them in the main list, and 54 which have been recorded as 

 Canarian birds from such unreliable sources that they may be rejected. 



The author's discussion of the origin and relationship of the Canarian 

 fauna and the problem of the origin of island faunas in general is full of 

 food for thought. He endorses the theory that the Canary Islands were 

 never part of the African mainland, their volcanic origin, deep water 

 separation, and absence of terrestrial mammals and reptiles being ample 

 evidence in the negative. The resident birds have therefore been de- 

 rived from migrants which have been stranded there and remained to 

 breed, and which have eventually become modified by the local environ- 

 ment. In this connection we find that 41 of the 61 resident forms are 

 of northern European affinities and all have closely related races in the 

 British Isles. 



The differentiation of races within the Canary group is particularly 

 interesting and as a rule we find one race of a species inhabiting the west- 

 ern group of islands and a different one in the far more arid eastern islands. 

 Here the peculiar desert environment has been active, as it has in produc- 

 ing the pale races of birds in the desert areas of western North America. 

 The distinct races of a few species, which we find inhabitingdifferentislands 

 in the western group, have been attributed by the author to successive 

 invasions of the migrating mainland birds at remote periods, but it seems 

 to us that this supposition is hardly necessary, since birds introduced 

 into two islands simultaneously may select a different sort of food on 

 each island even though the range of choice may be exactly the same, and 

 make other selections which in course of time would be reflected in their 

 color or size. Then too environments which may appear to us precisely 

 similar may have elements of difference that will have a marked effect 

 upon the birds that are brought under their influence. The most inter- 

 esting of the endemic birds of the Canaries are the two forms of the blue 

 Chaffinch (Fringilla teydea) which are found in the pine belts of the 

 high mountains of Tenerife and Gran Canaria, the low grounds of which 

 islands are inhabited by a form of Fringilla coelebs. These birds have no 

 close relative anywhere and are probably the oldest species of the endemic 

 avifauna. Mr. Bannerman suggests that an ancestral or allied species 

 might be logically looked for somewhere in the Atlas mountains of north- 

 ern Africa. It is inconceivable that such strikingly different birds could 



